Initial Design rules Stainless Steel 3D printing

This post is meant to give you guys a better indication of what is and what is not possible with 3D printing Stainless Steel. These guidelines are based on all your uploads and are a work in progress. They are meant to be indicative. We have images to illustrate these examples and I will add them once we have permission from the model owners.

I know that mass and weight are weird things to consider when 3D
modeling(unless you’re a CAD person of course). But, look at your model
and think, if I made this out of clay would it work?
This is really the best tip I can give you.

small parts up to 50 by 50mm

95% of these parts print successfully. They can have very thin walls of below 3 mm. But, the most important reason for failing is that the parts are not strong enough or do not self-support. The model will either fall apart or crush itself. Lets imagine we are trying to print a wine glass standing upright. A 3 mm thick stem might work if it were printed alone. But in a wine glass the stem has to support the bowl. During production the model is very fragile. If the bowl is heavy it will simply topple over.

Lets imagine now that we are trying to print this same wineglass horizontally. The stem would break in the middle. Lets now imagine that we are trying to print a wine glass that has no stem but simply consists of the bowl. This would work because a wine glass bowl is a self supporting structure without a weak point that could cause the model to break. What if we get ambitious and would like to add lots of decoration to our wineglass bowl? This is possible. But, if the decoration would consist of a lot of mass the weight would break the bowl during the printing process.

With Stainless Steel it is not a case of absolute wall thickness as with the other materials. The EOS printer that makes White, Strong & Flexible simply can not print thinner walls than 0.7mm (really, please only use 1mm or higher). With Stainless Steel there are a lot of variables. This is why it is so hard for us to give you guidance on what can and can not be done.

Very thin or delicate structures can not be printed either.

If it looks wispy and lovely and feather-like it will probably have an issue. People have to remove support material, put it in an oven etc. 

Integrated ‘whole’ parts have the highest chance of success. Lets imagine you wanted to print a plate of spaghetti. If it was one mass of spaghetti on a plate it might be possible. If there were strands of spaghetti that magically would stand upright and horizontally in all directions it would not work, it would break. If it would be and incredibly thin structure consisting of many individual strands in a bowl, it would not work. It could simply not be cleaned.

Furthermore since there is an oven step involved whereby the models are all heated to a high temperature parts might fuse if they are loose or close together. So, a bowl of many individual strands of spaghetti would fuse.

Medium to large parts less than 200 by 200 by 100mm

These parts need wall thicknesses of 3mm. Why? The part is larger so the internal structures supporting the part need to be stronger.

These are what I can give you right now. I hope to get more information to you as we learn.

Images are used under a Creative Commons Share, remix, attribution license. Images made by: KARPOV, AndrewK100 and Sebastian_Mary respectively.

Pin It


    1. Joris


      they’re great! A lot of fun!


  1. Madox.NET

    Propeller Hats!
    A picture post showing the latest delivery from Shapeways.Β  Joris just posted guidelines on stainless steel printing so I thought I’ll post some photos to show what tiny detail can be achieved.

    Fits quite nicely on the shirt cuf…

  2. Magic

    Very interesting the fact that for small object you can have wall of 1mm.
    I still do not understand if we can print moving parts: I am afraid that if there is no more support material during the infusion phase, the moving parts could get fused.
    For example could the Frankenstein Ring be printed in Stainless Steel ?
    Here is the model:

    1. Joris


      the tolerances and the accuracy would make fine moving parts difficult. Exact gears and things would seem to be difficult also. But, a moving part with a large gap should work.


    2. Magic

      Currently, this model has 1 mm of clearance and 2 mm of thickness. But I have plans to reduce the model of 75% (0.75 mm or clearance and 1.5 mm of thickness) and 50% (0,5 mm of cleareance and 1 mm of thickness) in White Strong and Flexible.
      I guess these reduced versions won’t work in Stainless Steel.
      I should also renounce to print a separate nut that could be screwed to the bolt… πŸ™

  3. George Bell

    What about printing that wine glass upside down? Would that work?

    Darn gravity. Maybe you can take a 3D printer up to the International Space Station? Then anything would work!

    1. Joris

      In terms of cost effectiveness we could do it in one of those Russian practice planes and do a layer every time the plane is at the top of its parabolic orbit.

      Does UPS even do pick ups from the ISS?

  4. Stijn van der Linden

    Thank you, this is very clarifying and helpful πŸ™‚
    Up till now it felt like modeling in the dark for SS, hoping and praying that the prints would succeed πŸ˜‰

    Could you also provide some info on the polishing step ?
    What limitations does this add ? Is there a minimum practical size for polishing ?
    With the braille and other rings I tested, it seems they were hardly polished at all, still very rough (almost like wsf) and red.
    I can understand some reluctance to polish fine details (for example; accidentally removing the braille dots πŸ˜‰ but there were a few rings that I think could have been a lot smoother and cleaner.

    I’ll post some pics as soon as the darn memory card will release them to me :-\

    1. Joris


      The polishing step is a manual one. So it is very variable. Polishing should not reduce overall surface detail nor significantly affect the size of your model.

      We’re working on improving this process but again it is a manual process so results may vary.

  5. Stephanie P

    So many people in London have skull rings, but this is my far the coolest.
    Very Damien Hirst πŸ˜‰

  6. DBucci

    What percentage of shrinkage occurs in post infiltrated parts?


  7. Glen Gardner

    To All,
    It first needs to be recognized that 3d metal printing is a technology that is in it’s infancy. Unlike traditional metal working methods with centuries long histories, many of the forces, stresses and reactions that printed metal parts must undergo have not been thoroughly quantified as hard data. Simply put, most of the time we know what works and what doesn’t, but we can’t always tell you exactly why. As in most technologies emerging from small companies, data collection is done on an “as needed” basis. The bulk of all human knowledge comes from data collected over years, decades or even centuries of trial and error. This might help explain why giving precise reasons and fool proof solutions for each and every failed or “non-printable” design is not practical or even possible.

    From my perspective here in the metal print shop I wish to offer some observations that I believe to be consistently true. When parts come out of the print machinery they consist of metal powder held together by a chemical binder (glue). All [I]green[/I] parts are fragile. Think in terms of air dried potters clay. As the size and intricacy of a design increases so does it’s fragility. In the green state, these parts are extremely difficult to process without breaking. If a part is able to be loaded into the sintering furnace a whole new set of forces now come into play. Observations inside the running furnace are impossible so finished parts can only be examined [I]after[/I] the furnace is cooled. The most common failures detected [I]after[/I] sintering take the form of hairline cracks or open breaks. This type of flaw usually occurs in areas consisting of sharp angular intersections and/or radical section changes. The use of fillets and smooth transitions can help prevent this from occurring but not in every case. Flat broad thin meshes and grids are very often problematic but not always impossible. The overall size to mass ratio does seem to play a factor but no generalized rule has yet to be determined. Actually, it would surprise me if the use of the true “Golden Ratio” (1.618033……) didn’t improve any design intended for 3dmp.

    Forgive me if these comments seem a bit vague but currently it is simply the nature of the beast. I truly believe that [I]everyone[/I] involved with these new technologies is now acting as a contributor to a global knowledge base. Whether designer, businessman or technician, we are all together writing the book on 3d printing/manufacturing. Pretty exciting stuff really!

    -Glen Gardner

  8. ideazonepd

    I’m curious to know if it would be possible to make a Glossy Stainless Steel material available on my models? (similar to the glossy on Gold Plated and Antique Bronze).



  9. bob scifo

    Stainless steel seems to shrink very slightly; is there a known percentage of shrinkage?

Comments are closed.